المنهج

عناوين الوحدات والدروس فقط. يُفتح المحتوى الكامل بعد التسجيل.

1. Module M01

   1. Lesson M01-L01 — Query rewriting for retrieval
   2. Lesson M01-L02 — Query decomposition and multi-query retrieval
   3. Lesson M01-L03 — Contextual and hierarchical retrieval
   4. Lesson M01-L04 — Hybrid retrieval, reranking, and retrieval routing

2. Module M02

   1. Lesson M02-L01 — Answer-with-evidence design
   2. Lesson M02-L02 — Citations and source attribution
   3. Lesson M02-L03 — Abstention, conflicts, and incomplete evidence
   4. Lesson M02-L04 — Trust boundaries and prompt injection from retrieved content

3. Module M03

   1. Lesson M03-L01 — Golden datasets for RAG evaluation
   2. Lesson M03-L02 — Retrieval metrics: precision, recall, ranking signals
   3. Lesson M03-L03 — Answer/groundedness evaluation and judge limitations
   4. Lesson M03-L04 — Failure taxonomy, regression suites, cost/latency eval

4. Module M04

   1. Lesson M04-L01 — When to retrieve vs not retrieve
   2. Lesson M04-L02 — Bounded iterative / tool-assisted retrieval
   3. Lesson M04-L03 — Observability, caching, failure recovery
   4. Lesson M04-L04 — Production readiness and ship JP-JA17